Improving plan about lifetime prediction of gas turbine throughout image processing

Many systems which have introduced Image Processing Technique are being used in various fields. In this paper, we predict the lifetime of a gas turbines as one of the image processing applications. A gas turbine, high-priced key module, is very useful for repairing and recycling throughout prediction of lifetime. Taking the place of lifetime predicted by existing expert´s naked eyes, remaining life can be estimated by processing digital images. The accuracy of lifetime prediction fell because precipitates wasn´t separated or extracted from existing research. Then, we propose compound image analyzing method to improve the accuracy. We also provide exact information for lifetime prediction of blade by combination techniques - separation of region from the original document, separation of contour line for getting an exact size, filling of region etc. The number of extracted precipitates from proposed method has been improved by 15~20% more than the existing one.